But our programs worked by creating virtual agents inside the computer. Julia had created real agents in the real world. At first I didn't see how our programs could be adapted to what she was doing.

"We use them for structure," she said. "The program makes the swarm structure." Of course. It was obvious that a single molecular camera was inadequate to register any sort of image. Therefore, the image must be a composite of millions of cameras, operating simultaneously. But the cameras would also have to be arranged in space in some orderly structure, probably a sphere. That was where the programming came in. But that in turn meant that Xymos must be generating the equivalent of"You're making an eye."

"Kind of. Yes."

"But where's the light source?"

"The bioluminescent perimeter."

"That's not enough light."

"It is. Watch."

Meanwhile, the onscreen Julia was turning smoothly, pointing to the intravenous line behind her. She lifted a syringe out of a nearby ice bucket. The barrel appeared to be filled with water. "This syringe," she said, "contains approximately twenty million cameras in isotonic saline suspension. At the moment they exist as particles. But once they are injected into the bloodstream, their temperature will increase, and they will soon flock together, and form a meta-shape. Just like a flock of birds forms a V-shape."

"What kind of a shape?" one of the VCs asked.

"A sphere," she said. "With a small opening at one end. You might think of it as the equivalent of a blastula in embryology. But in effect the particles form an eye. And the image from that eye will be a composite of millions of photon detectors. Just as the human eye creates an image from its rods and cone cells."